Help with a new study ranking domain companies by how healthy they are.
The domain name industry is fixated on domains under management, or DUMs.
We tend to rank registrars by how many domains their customers have registered and consider registries with the most registrations to be more successful.
But the raw numbers are just part of the story.
Registrars can sell cheap domains in bulk. Registries can stuff free domains into the registrants’ hands.
Some registrars and registries are “healthier” than others. I’m seeking to quantify which registrars and registries have the healthiest businesses in a study in conjunction with DataProvider.com. We’re digging below the topline numbers to consider how domains are used.
While we have plenty of ideas on how to measure this, I’m curious what Domain Name Wire readers think. If you have thoughts on what makes a registrar or registry healthy, please email me at andrew (at) domain name wire .com, or drop a comment below or on social media.
As with all tech-driven companies, a healthy registrar or registry is one that can implement and maintain its own technology effectively, without accumulating significant technical debt.
On the registry side, one indicator is whether orders go through automatically or require support requests and manual attention — so on this basis, the .NYC, .PK and the post-GMO .SO registries are not healthy.
For a registry, do they confiscate valuable domains without due process or transparency? Then what happens to confiscated domains? I’m particularly looking at .CN and .IN/.BHARAT here.
For registrars, one can look at cart or order-failure rates, although order failures for the less popular TLD are frequently if not entirely due to registries not being fully or properly automated.
A good example of technical debt at the registrar level is what has become of the BR/enom integration at Tucows, where actual domain inventories don’t always match what customers see in their account control panels, and changes in DNS settings must sometimes be repeated before they take effect. Fortunately, Tucows has a history of hiring exceptionally competent support staff.
Registrars’ and registries’ TOSs are another indication of health. Does an entity follow their own TOS? Do they refund customers’ funds if a domain registration doesn’t go through, especially during land rushes? Does a TOS provide any protection for registrants if a government agency or private party anywhere in the world complains about a domain being registered or how it’s being used?
Last but not least, how healthy and safe is it for staff working at a registrar or registry? How healthy is the atmosphere and ventilation? Are there smoke detectors and fire suppression systems? Are children employed and if so, what protections are in place for them? Is a workforce roughly representative of the local population?
It is more than simply how domain names are used and Web Usage is only part of the equation. Many domain names are not used for websites. While digging below the raw numbers is good, there are many registry and registrar models. I have compiled the ICANN registrar data on a per gTLD and per registrar basis back to 2001 when ICANN started publishing that data. ICANN only adopted a more uniform reporting model in 2014. Over that time, there have been significant changes in registry and registrar models. It is also possible to see year on year, five year, ten year, fifteen and twenty year changes on the registrars.
The old ICANN model was that of Registry-Registrar-Registrant (3R). It was simple enough. However, the costs of becoming an ICANN accredited registrar are high compared to the costs of becoming an accredited registrar in the local ccTLD. The rise of the ccTLDs has also been a major change as when the local ccTLD begins to dominate a country level market, the new registration volume in the gTLDs begins to fall. The core legacy gTLDs (.COM/NET/ORG) still make up the main gTLD registration in these countries each month. The gTLD registrations tend to be either on the few local ICANN registrars or outsourced to ICANN registrars. In effect this creates another layer for the old ICANN 3R model and adds the Reseller level. I crunch this data each month for some of the reports that I publish and there is even a seasonality to the transactions on registrars and resellers. Some of that is down to registries and registrars using discounting to drive registrations.
The reseller level is much more complex than the well documented 3R model. Approximately 21% of the gTLD market is not hosted on registrar nameservers. That’s only part of the reseller market. The other part of the reseller market consists of resellers that use the registrar nameservers while maintaining their own web hosting. Some of the larger registrars have thousands of such customers. Again, some registrar business models are not focused on retail domain name sales. Others use domain names as a loss leader to upsell customers to products with a higher profit margin. The bulk of ICANN registrars are not retail-facing registrars. They are drop catcher registrars.
The problem with the registrar data supplied in the ICANN registry reports is that it is summary data. This makes it impossible, or very difficult, to derive some of the most important metrics of what makes a registry or registrar healthy. Those metrics go well beyond usage and are dependent on the business model used by the registry or registrar. And some gTLDs are really two TLDs in one.
In my opinion NOT having cloud flare would be a start!!!
Cloudflare is an ICANN accredited registrar and it also provides CDN and DNS services. The starting point for a study would be the list of ICANN accredited registrars. It has about 1.9 million gTLD domain names as a registrar. It provides DNS for about 20 million.
.in registry taken away bank.in fin.in from registrants forcibly because after 20 years the fools at RBI wants to use those domains as brands. So a registry act like a scumbug which enjoy stealing users domain by showing bad faith